U.S. patent number 8,761,312 [Application Number 11/351,683] was granted by the patent office on 2014-06-24 for selection of a thresholding parameter for channel estimation.
This patent grant is currently assigned to Qualcomm Incorporated. The grantee listed for this patent is Durga Prasad Malladi, Anastasios Stamoulis. Invention is credited to Durga Prasad Malladi, Anastasios Stamoulis.
United States Patent |
8,761,312 |
Stamoulis , et al. |
June 24, 2014 |
Selection of a thresholding parameter for channel estimation
Abstract
Techniques for deriving a high quality channel estimate are
described. A first channel impulse response estimate (CIRE) having
multiple channel taps is derived, e.g., by filtering initial CIREs
obtained from a received pilot. A threshold parameter value is
selected based on at least one criterion, which may relate to
channel profile, operating SNR, number of channel taps, and so on.
A second CIRE is derived by zeroing out selected ones of the
channel taps in the first CIRE based on the threshold parameter
value. The average energy of the channel taps may be determined, a
threshold may be derived based on the average energy and the
threshold parameter value, and channel taps with energy less than
the threshold may be zeroed out. A memory may store threshold
parameter values for different operating scenarios, and a stored
value may be selected for use based on the current operating
scenario.
Inventors: |
Stamoulis; Anastasios (San
Diego, CA), Malladi; Durga Prasad (San Diego, CA) |
Applicant: |
Name |
City |
State |
Country |
Type |
Stamoulis; Anastasios
Malladi; Durga Prasad |
San Diego
San Diego |
CA
CA |
US
US |
|
|
Assignee: |
Qualcomm Incorporated (San
Diego, CA)
|
Family
ID: |
36501922 |
Appl.
No.: |
11/351,683 |
Filed: |
February 10, 2006 |
Prior Publication Data
|
|
|
|
Document
Identifier |
Publication Date |
|
US 20060227748 A1 |
Oct 12, 2006 |
|
Related U.S. Patent Documents
|
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
Issue Date |
|
|
60652236 |
Feb 11, 2005 |
|
|
|
|
Current U.S.
Class: |
375/340 |
Current CPC
Class: |
H04L
25/0228 (20130101); H04L 25/0258 (20130101); H04L
25/0256 (20130101); H04L 27/2647 (20130101); H04L
25/0212 (20130101); H04L 25/0204 (20130101); H04L
25/025 (20130101); H04L 27/261 (20130101) |
Current International
Class: |
H04L
27/06 (20060101) |
Field of
Search: |
;375/142,143,144,148,150,152,316,343,346 ;455/63.1,114.2,278.1,296
;370/332,322 |
References Cited
[Referenced By]
U.S. Patent Documents
Foreign Patent Documents
|
|
|
|
|
|
|
0966113 |
|
Dec 1999 |
|
EP |
|
2000091968 |
|
Mar 2000 |
|
JP |
|
2002252575 |
|
Sep 2002 |
|
JP |
|
2004104225 |
|
Apr 2004 |
|
JP |
|
2004266814 |
|
Sep 2004 |
|
JP |
|
2005051404 |
|
Feb 2005 |
|
JP |
|
2006262039 |
|
Sep 2006 |
|
JP |
|
2006295293 |
|
Oct 2006 |
|
JP |
|
2144733 |
|
Jan 2000 |
|
RU |
|
2157053 |
|
Sep 2000 |
|
RU |
|
2214686 |
|
Oct 2003 |
|
RU |
|
1225023 |
|
Apr 1986 |
|
SU |
|
0145296 |
|
Jun 2001 |
|
WO |
|
WO0209297 |
|
Jan 2002 |
|
WO |
|
02082683 |
|
Oct 2002 |
|
WO |
|
Other References
International Search Report--PCT/US2006/005185, International
Search Authority--European Patent Office--Jun. 19, 2006. cited by
applicant .
International Preliminary Examination Report--PCT/US2006/005185,
International Search Authority--European Patent Office--Aug. 14,
2007. cited by applicant .
Written Opinion--PCT/US2006/005185, International Search
Authority--European Patent Office--Aug. 11, 2007. cited by
applicant .
Taiwan Search Report--TW095104643--TIPO--Aug. 30, 2012. cited by
applicant.
|
Primary Examiner: Tran; Khanh C
Attorney, Agent or Firm: Katbab; Abdollah
Parent Case Text
CLAIM OF PRIORITY UNDER 35 U.S.C. .sctn.119
The present Application for Patent claims priority to Provisional
Application Ser. No. 60/652,236, entitled "SELECTION OF A
THRESHOLDING PARAMETER FOR CHANNEL ESTIMATION," filed Feb. 11,
2005, assigned to the assignee hereof, and expressly incorporated
herein by reference.
Claims
What is claimed is:
1. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, a second CIRE by zeroing out a percentage of
the multiple channel taps based on a threshold parameter value; the
at least one processor configured to determine a new threshold
parameter value based on the second CIRE and zeroing out a further
percentage of the multiple channel taps based on the new threshold
parameter value; and a memory coupled to the at least one
processor.
2. The apparatus of claim 1, wherein the at least one processor
derives a threshold based on the multiple channel taps and the
threshold parameter value.
3. The apparatus of claim 1, wherein the at least one processor
determines an average energy of the multiple channel taps, derives
a threshold based on the average energy and the threshold parameter
value, and derives the second CIRE by zeroing out channel taps with
energy less than the threshold.
4. The apparatus of claim 1, wherein the at least one processor is
configured to determine whether the second CIRE channel estimate
improvement is desired prior to determining the new threshold
parameter value.
5. The apparatus of claim 4, wherein improvement of the second CIRE
is desired when a packet is decoded in error.
6. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, select a threshold parameter value based on
at least one criterion, derive a second CIRE by zeroing out
selected ones of the multiple channel taps based on the threshold
parameter value, ascertain whether a packet is decoded in error
and, if the packet is decoded in error, select a new threshold
parameter value and derive a new second CIRE by zeroing out
selected ones of the multiple channel taps based on the new
threshold parameter value; and a memory coupled to the at least one
processor.
7. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, ascertain an operating signal-to-noise ratio
(SNR) of the channel, to select a threshold parameter value based
on the operating SNR, and to derive a second CIRE by zeroing out
selected ones of the multiple channel taps based on the threshold
parameter value; and a memory coupled to the at least one
processor; wherein the at least one processor selects different
threshold parameter values and derives different second CIREs based
on the first CIRE and the different threshold parameter values
until a packet is decoded correctly or a termination condition is
encountered.
8. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, to select a threshold parameter value, and
to derive a second CIRE by zeroing out selected ones of the
multiple channel taps based on the threshold parameter value; and a
memory coupled to the at least one processor; wherein the memory
stores a table of threshold parameter values for different
operating scenarios, and wherein the at least one processor is
configured to select the threshold parameter value from the table
based on current operating scenario that includes coding and
modulation scheme of the apparatus.
9. The apparatus of claim 8, wherein the current operating scenario
comprises a channel profile.
10. The apparatus of claim 9, wherein the channel profile further
comprises at least one signal delay and at least one signal
power.
11. An apparatus comprising: at least one processor configured to
derive initial channel impulse response estimates (CIREs) for
multiple symbol periods based on received pilot and derive a first
CIRE having multiple channel taps by filtering the initial CIREs,
to select a threshold parameter value based on at least one
criterion, and to derive a second CIRE by zeroing out selected ones
of the multiple channel taps based on the threshold parameter
value; and a memory coupled to the at least one processor.
12. The apparatus of claim 11, wherein the multiple symbol periods
comprise a current symbol period, at least one prior symbol period,
and at least one future symbol period.
13. The apparatus of claim 11, wherein the multiple symbol periods
comprise a current symbol period and at least one prior symbol
period.
14. The apparatus of claim 11, wherein the multiple symbol periods
comprise a current symbol period and at least one future symbol
period.
15. A method of selecting a threshold parameter for channel
estimating, comprising: deriving a first channel impulse response
estimate (CIRE) having multiple channel taps; selecting a threshold
parameter value based on at least one criterion; and deriving a
second CIRE by zeroing out selected ones-of the multiple channel
taps based on the threshold parameter value; wherein the selecting
the threshold parameter value comprises selecting the threshold
parameter value based on channel profile that includes at least one
signal delay value to an apparatus, wherein at least one of the
above acts is performed by a processor; ascertaining whether a
packet is decoded in error; and if the packet is decoded in error,
selecting a new threshold parameter value, and deriving a new
second CIRE by zeroing out selected ones of the multiple channel
taps based on the new threshold parameter value.
16. A method of selecting a threshold parameter for channel
estimating, comprising: deriving a first channel impulse response
estimate (CIRE) having multiple channel taps; selecting a threshold
parameter value based on at least one criterion; deriving a second
CIRE by zeroing out selected ones of the multiple channel taps
based on the threshold parameter value; and deriving initial CIREs
for multiple symbol periods based on received pilot; wherein the
deriving the first CIRE comprises filtering the initial CIREs to
obtain the first CIRE, wherein at least one of the above acts is
performed by a processor.
17. An apparatus comprising: means for deriving a first channel
impulse response estimate (CIRE) having multiple channel taps;
means for selecting a threshold parameter value based on at least
one criterion; and means for deriving a second CIRE by zeroing out
a percentage of the multiple channel taps based on the threshold
parameter value; wherein the means for selecting the threshold
parameter value comprises means for selecting the threshold
parameter value based on channel profile that includes at least one
signal delay value to the apparatus; means for ascertaining whether
a packet is decoded in error; means for selecting a new threshold
parameter value if the packet is decoded in error; and means for
deriving a new second CIRE, if the packet is decoded in error, by
zeroing out selected ones of the multiple channel taps based on the
new threshold parameter value.
18. An apparatus comprising: means for deriving a first channel
impulse response estimate (CIRE) having multiple channel taps;
means for selecting a threshold parameter value based on at least
one criterion; means for deriving a second CIRE by zeroing out
selected ones of the multiple channel taps based on the threshold
parameter value; and means for deriving initial CIREs for multiple
symbol periods based on received pilot; wherein the means for
deriving the first CIRE comprises means for filtering the initial
CIREs to obtain the first CIRE.
19. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, to select a threshold parameter value based
on at least one criterion, wherein the at least one criterion
comprises a coding rate of received data, and to derive a second
CIRE by zeroing out selected ones of the multiple channel taps
based on the threshold parameter value; and a memory coupled to the
at least one processor.
20. The apparatus of claim 19, wherein the threshold parameter
value varies inversely with the coding rate.
21. An apparatus comprising: at least one processor configured to
derive a first channel impulse response estimate (CIRE) having
multiple channel taps, to select a threshold parameter value based
on at least one criterion, wherein the at least one criterion
comprises a modulation scheme of received data, and to derive a
second CIRE by zeroing out selected ones of the multiple channel
taps based on the threshold parameter value; and a memory coupled
to the at least one processor.
22. The apparatus of claim 21, wherein the threshold parameter
value varies inversely with a number of constellation points of the
modulation scheme.
Description
BACKGROUND
I. Field
The present disclosure relates generally to communication, and more
specifically to techniques for deriving a channel estimate for a
communication channel.
II. Background
In a communication system, a transmitter typically processes (e.g.,
encodes, interleaves, and symbol maps) traffic data to generate
data symbols, which are modulation symbols for data. For a coherent
system, the transmitter multiplexes pilot symbols with the data
symbols, processes the multiplexed data and pilot symbols to
generate a radio frequency (RF) signal, and transmits the RF signal
via a communication channel. The channel distorts the RF signal
with a channel response and further degrades the RF signal with
noise and interference.
A receiver receives the transmitted RF signal and processes the
received RF signal to obtain samples. For coherent data detection,
the receiver estimates the response of the communication channel
based on the received pilot and derives a channel estimate. The
receiver then performs data detection (e.g., equalization) on the
samples with the channel estimate to obtain data symbol estimates,
which are estimates of the data symbols sent by the transmitter.
The receiver then processes (e.g., demodulates, deinterleaves, and
decodes) the data symbol estimates to obtain decoded data.
The quality of the channel estimate may have a large impact on data
detection performance and may affect the quality of the symbol
estimates as well as the correctness of the decoded data. There is
therefore a need in the art for techniques to derive a high quality
channel estimate in a communication system.
SUMMARY
Techniques for deriving a high quality channel estimate are
described herein. According to an embodiment of the invention, an
apparatus is described which includes at least one processor and a
memory. The processor(s) derive a first channel impulse response
estimate (CIRE) having multiple channel taps. The processor(s) may
derive initial CIREs based on a received pilot and may filter the
initial CIREs to obtain the first CIRE. The processor(s) select a
threshold parameter value based on at least one criterion, which
may relate to channel profile, operating signal-to-noise ratio
(SNR), expected channel delay spread, number of channel taps, and
so on. The processor(s) derive a second CIRE by zeroing out
selected ones of the channel taps in the first CIRE based on the
threshold parameter value. The processor(s) may determine the
average energy of the channel taps, derive a threshold based on the
average energy and the threshold parameter value, and zero out
channel taps with energy less than the threshold. The memory may
store a table of threshold parameter values for different operating
scenarios. The processor(s) may select one of the stored threshold
parameter values based on the current operating scenario.
According to another embodiment, a method is provided in which a
first CIRE having multiple channel taps is derived. A threshold
parameter value is selected based on at least one criterion. A
second CIRE is derived by zeroing out selected ones of the multiple
channel taps based on the threshold parameter value.
According to yet another embodiment, an apparatus is described
which includes means for deriving a first CIRE having multiple
channel taps, means for selecting a threshold parameter value based
on at least one criterion, and means for deriving a second CIRE by
zeroing out selected ones of the multiple channel taps based on the
threshold parameter value.
Various aspects and embodiments of the invention are described in
further detail below.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows a block diagram of a transmitter and a receiver.
FIG. 2 shows an exemplary multi-tier frame structure.
FIG. 3 shows an exemplary subband structure.
FIG. 4 illustrates thresholding for a channel impulse response
estimate.
FIG. 5 shows a block diagram of a channel estimator/processor at
the receiver.
FIG. 6 shows a process for performing channel estimation with
thresholding.
DETAILED DESCRIPTION
The word "exemplary" is used herein to mean "serving as an example,
instance, or illustration." Any embodiment described herein as
"exemplary" is not necessarily to be construed as preferred or
advantageous over other embodiments.
The channel estimation techniques described herein may be used for
various communication systems such as a Code Division Multiple
Access (CDMA) system, a Time Division Multiple Access (TDMA)
system, a Frequency Division Multiple Access (FDMA) system, an
orthogonal frequency division multiplexing (OFDM) system, an
orthogonal frequency division multiple access (OFDMA) system, a
single-carrier FDMA (SC-FDMA) system, and so on. A CDMA system may
implement one or more radio access technologies (RATs) such as
Wideband-CDMA (W-CDMA), cdma2000, and so on. cdma2000 covers
IS-2000, IS-856, and IS-95 standards. A TDMA system may implement a
RAT such as Global System for Mobile Communications (GSM). These
various RATs and standards are known in the art. An OFDM system may
be an IEEE 802.11a/g system, a Digital Video Broadcasting for
Handhelds (DVB-H) system, an Integrated Services Digital
Broadcasting for Terrestrial Television Broadcasting (ISDB-T)
system, and so on. An OFDMA system transmits modulation symbols in
the frequency domain on orthogonal frequency subbands using OFDM.
An SC-FDMA system transmits modulation symbols in the time domain
on orthogonal frequency subbands. For clarity, the techniques are
described below for a system with multiple frequency subbands,
which may be an OFDM, OFDMA, or SC-FDMA system.
FIG. 1 shows a block diagram of a transmitter 110 and a receiver
150 in a wireless communication system 100. For simplicity,
transmitter 110 and receiver 150 are each equipped with a single
antenna. For the downlink (or forward link), transmitter 110 may be
part of a base station, and receiver 150 may be part of a terminal.
For the uplink (or reverse link), transmitter 110 may be part of a
terminal, and receiver 150 may be part of a base station. A base
station is typically a fixed station and may also be called a base
transceiver system (BTS), an access point, a Node B, or some other
terminology. A terminal may be fixed or mobile and may be a
wireless device, a cellular phone, a personal digital assistant
(PDA), a wireless modem card, and so on. The channel estimation
techniques described herein may be used for a terminal as well as a
base station.
At transmitter 110, a transmit (TX) data processor 112 processes
(e.g., encodes, interleaves, and symbol maps) traffic data and
generates data symbols. A pilot processor 114 generates pilot
symbols. As used herein, a data symbol is a modulation symbol for
data, a pilot symbol is a modulation symbol for pilot, a modulation
symbol is a complex value for a point in a signal constellation
(e.g., for PSK or QAM), and a symbol is generally a complex value.
A modulator 120 multiplexes the data symbols and pilot symbols,
performs modulation (e.g., for OFDM or SC-FDMA) on the multiplexed
data and pilot symbols, and generates transmission symbols. A
transmission symbol may be an OFDM symbol or an SC-FDMA symbol and
is sent in one symbol period. A transmitter unit (TMTR) 132
processes (e.g., converts to analog, amplifies, filters, and
frequency upconverts) the transmission symbols and generates an RF
signal, which is transmitted via an antenna 134.
At receiver 150, an antenna 152 receives the RF signal from
transmitter 110 and provides a received signal to a receiver unit
(RCVR) 154. Receiver unit 154 conditions (e.g., filters, amplifies,
frequency downconverts, and digitizes) the received signal and
provides input samples. A demodulator 160 performs demodulation
(e.g., for OFDM or SC-FDMA) on the input samples to obtain received
symbols. Demodulator 160 provides received pilot symbols to a
channel estimator/processor 170 and provides received data symbols
to a data detector 172. Channel estimator/processor 170 derives
channel estimates for the wireless channel between transmitter 110
and receiver 150 based on the received pilot symbols. Data detector
172 performs data detection (e.g., equalization or matched
filtering) on the received data symbols with the channel estimates
and provides data symbol estimates, which are estimates of the data
symbols sent by transmitter 110. An RX data processor 180 processes
(e.g., symbol demaps, deinterleaves, and decodes) the data symbol
estimates and provides decoded data. In general, the processing at
receiver 150 is complementary to the processing at transmitter
110.
Controllers/processors 140 and 190 direct the operation of various
processing units at transmitter 110 and receiver 150, respectively.
Memories 142 and 192 store program codes and data for transmitter
110 and receiver 150, respectively.
FIG. 2 shows an exemplary multi-tier frame structure 200 that may
be used for system 100. The transmission time line is partitioned
into super-frames, with each super-frame having a predetermined
time duration, e.g., approximately one second. Each super-frame may
include (1) a header field for a time division multiplexed (TDM)
pilot and overhead/control information and (2) a data field for
traffic data and a frequency division multiplexed (FDM) pilot. The
data field may be partitioned into multiple (O) equal-size
outer-frames, each outer-frame may be partitioned into multiple (F)
frames, and each frame may be partitioned into multiple (T) slots.
For example, each super-frame may include four outer-frames (O=4),
each outer-frame may include 32 frames (F=32), and each frame may
include 15 time slots (T=15). If each frame has a duration of 10
milliseconds (ms), which conforms to W-CDMA, then each slot has a
duration of 667 microseconds (.mu.s), each outer-frame has a
duration of 320 ms, and each super-frame has a duration of
approximately 1.28 seconds. The super-frame, outer-frame, frame,
and slot may also be referred to by some other terminology.
In an embodiment, different radio technologies may be used for
different slots. For example, W-CDMA may be used for some slots,
and OFDM may be used for other slots. In general, the system may
support any one or any combination of radio technologies, and each
slot may employ one or multiple radio technologies. A slot used for
OFDM is called an OFDM slot. An OFDM slot may carry one or more (N)
OFDM symbols and may further include a guard period (GP). For
example, an OFDM slot may carry three OFDM symbols and a guard
period, with each OFDM symbol having a duration of approximately
210 .mu.s.
FIG. 3 shows an exemplary subband structure 300 that may be used
for system 100. The system has an overall system bandwidth of BW
MHz, which is partitioned into multiple (K) orthogonal subbands. K
may be any integer value but is typically a power of two (e.g.,
128, 256, 512, 1024, and so on) in order to simplify the
transformation between time and frequency. The spacing between
adjacent subbands is BW/K MHz. In a spectrally shaped system, G
subbands are not used for transmission and serve as guard subbands
to allow the system to meet spectral mask requirements, where
typically G>1. The G guard subbands are often distributed such
that G.sub.L.apprxeq.G/2 guard subbands are at the lower band edge
and G.sub.U.apprxeq.G/2 guard subbands are at the upper band edge.
The remaining U=K-G subbands may be used for transmission and are
called usable subbands.
To facilitate channel estimation, a pilot may be transmitted on a
set of M subbands that may be uniformly distributed across the
entire system bandwidth.
Consecutive subbands in the set may be spaced apart by S subbands,
where S=K/M.
Some of the subbands in the set may be among the G.sub.L lower
guard subbands and would not be used for pilot transmission, and
some other subbands in the set may be among the G.sub.U upper guard
subbands and would also not be used for pilot transmission. For the
example shown in FIG. 2, the first Z.sub.L subbands in the set are
not used for pilot transmission and are called zeroed-out pilot
subbands, the next P subbands in the set are used for pilot
transmission and are called used pilot subbands, and the last
Z.sub.U subbands in the set are zeroed-out pilot subbands, where
M=Z.sub.L+P+Z.sub.U.
In one exemplary design, the system utilizes a subband structure
with K=1024 total subbands, G.sub.L=68 lower guard subbands,
G.sub.U=68 upper guard subbands, U=888 usable subbands, M=128 pilot
subbands, P=111 usable pilot subbands, and C=108 chips for a cyclic
prefix appended to each OFDM symbol. Other values may also be used
for these parameters.
FIG. 2 shows an exemplary frame structure, and FIG. 3 shows an
exemplary subband structure. The channel estimation techniques
described herein may be used with various frame and subband
structures.
For clarity, the following nomenclature is used in the description
below. Vectors are denoted by holded and underlined texts with
subscript indicating the vector length, e.g., h.sub.M for an
M.times.1 vector or H.sub.K for a K.times.1 vector, where the
".times.1" in the dimension is implicit and omitted for clarity.
Matrices are denoted by bolded and underlined texts with subscript
indicating the matrix dimension, e.g., W.sub.M.times.K for an
M.times.K matrix. Time-domain vectors are generally denoted with
lower case texts, e.g., h.sub.K, and frequency-domain vectors are
generally denoted with upper case texts, e.g., H.sub.K.
The wireless channel between transmitter 110 and receiver 150 may
be characterized by either a time-domain channel impulse response
h.sub.K or a corresponding frequency-domain channel frequency
response H.sub.K. The relationships between the channel impulse
response and the channel frequency response may be expressed in
matrix form as follows: H.sub.K=W.sub.K.times.Kh.sub.K, and Eq (1)
h.sub.K=W.sub.K.times.K.sup.-1H.sub.K, Eq (2) where h.sub.K is a
K.times.1 vector for the impulse response of the wireless
channel,
H.sub.K is a K.times.1 vector for the frequency response of the
wireless channel,
W.sub.K.times.K is a K.times.K Fourier matrix,
.times..times. ##EQU00001## is a K.times.K inverse Fourier matrix,
and
".sup.H" denotes a conjugate transpose.
Equation (1) indicates that the channel frequency response is the
fast Fourier transform or discrete Fourier transform (FFT/DFT) of
the channel impulse response. Equation (2) indicates that the
channel impulse response is the inverse FFT or inverse DFT
(IFFT/IDFT) of the channel frequency response. The element in row r
and column c of the Fourier matrix W.sub.K.times.K may be given
as:
e.pi..times. ##EQU00002## for r=1, . . . , K and c=1, . . . , K. Eq
(3)
The "-1" in the exponent in equation (3) is due to indices r and c
starting with 1 instead of 0.
Transmitter 110 transmits data and pilot symbols on the usable
subbands to receiver 150. The data and pilot symbols may be assumed
to have an average energy of E.sub.s, or E{|X(k)|.sup.2}=E.sub.s,
where X(k) is a symbol transmitted on subband k and E{ } denotes an
expectation operation. For simplicity, the following description
assumes that each symbol is transmitted at unit power so that
E.sub.s=1.
The received symbols obtained by receiver 150 in OFDM symbol period
n may be expressed as:
Y.sub.K(n)=H.sub.K(n).smallcircle.X.sub.K(n)+.eta..sub.K(n). Eq (4)
where X.sub.K(n) is a K.times.1 vector containing the transmitted
symbols for the K subbands,
Y.sub.K(n) is a K.times.1 vector containing the received symbols
for the K subbands,
.eta..sub.K(n) is a K.times.1 vector of noise for the K subbands,
and
".smallcircle." denotes an element-wise product.
Each entry of X.sub.K(n) may be a data symbol for a data subband, a
pilot symbol for a pilot subband, or a zero symbol for an unused
subband (e.g., a guard subband). For simplicity, the pilot symbols
may be assumed to have a complex value of 1+j0 and a magnitude of
{square root over (E.sub.s)}=1. In this case, the received pilot
symbols are simply noisy versions of the channel gains in
H.sub.K(n).
If only P pilot subbands are used for pilot transmission, as shown
in FIG. 3, then the receiver may form an M.times.1 vector
Y.sub.M(n) containing P received pilot symbols for the P used pilot
subbands and Z.sub.L+Z.sub.U zero symbols for the zeroed-out pilot
subbands. Vector Y.sub.M(n) may be expressed as:
.function.--.function.--.times..times. ##EQU00003## where
0.sub.Z.sub.L and 0.sub.Z.sub.U are vectors of all zeros, and
Y.sub.P(n) is a P.times.1 vector of received pilot symbols for the
P used pilot subbands.
Various techniques may be used to estimate the channel impulse
response based on the received pilot symbols. These techniques
include a least-squares (LS) technique, a minimum mean square error
(MMSE) technique, a robust MMSE technique, and a zero-forcing (ZF)
technique.
A least-squares channel impulse response estimate (CIRE)
h.sub.M.sup.ls(n) may be derived as:
.function..times..times..function..times..times..times..function..eta..f-
unction..times..function..times..eta..function..times..times.
##EQU00004## where h.sub.M(n) is an M.times.1 channel impulse
response vector with M channel taps, and
.eta..sub.M(n) is an M.times.1 vector of noise for the M pilot
subbands.
Equation (6) indicates that the least-squares CIRE may be obtained
by simply taking an M-point IFFT/IDFT of the received pilot symbols
in Y.sub.M(n). A zero-forcing CIRE is equal to the least-squares
CIRE.
An MMSE CIRE h.sub.M.sup.mmse(n) may be derived as:
h.sub.M.sup.mmse(n)=.PSI..sub.hhW.sub.M.times.M.sup.-1[W.sub.M.times.M.PS-
I..sub.hhW.sub.M.times.M.sup.-1+.LAMBDA..sub..eta..eta.].sup.-1Y.sub.M(n),
Eq (7) where .PSI..sub.hh=E{h.sub.M(n)h.sub.M.sup.H(n)} is an
M.times.M channel covariance matrix, and
.LAMBDA..sub..eta..eta.=E{.eta..sub.M(n).eta..sub.M.sup.H(n)} is an
M.times.M noise covariance matrix.
A robust MMSE CIRE h.sub.M.sup.rmmse(n) may be derived as:
.function..times..function..times..times. ##EQU00005## Equation (8)
assumes that the taps in the channel impulse response are
uncorrelated and have equal power, so that
.PSI..sub.hh=I.sub.M.times.M. Equation (8) further assumes that the
noise .eta..sub.M(n) is additive white Gaussian noise (AWGN) with a
zero mean vector and a covariance matrix of
.LAMBDA..sub..eta..eta.=N.sub.0I.sub.K.times.K, where N.sub.0 is
the variance of the noise and I.sub.K.times.K is a K.times.K
identity matrix.
The receiver may derive an initial CIRE h'.sub.M(n) for each OFDM
symbol period n with pilot transmission based on received pilot
symbols from the OFDM symbol sent in that symbol period. The
receiver may derive h'.sub.M(n) using the least-squares, MMSE,
robust MMSE, or some other technique. Hence, h'.sub.M(n) may be
equal to h.sub.M.sup.ls(n), h.sub.M.sup.mmse(n) or
h.sub.M.sup.rmmse(n).
The receiver may filter the initial CIREs h'.sub.M(n) for different
OFDM symbol periods to obtain filtered CREs {tilde over
(h)}.sub.M(n) having improved quality. The filtering may be
performed in various manners.
In an embodiment, the filtering for an "interior" OFDM symbol n
that is bordered by OFDM symbol n-1 on the left side and OFDM
symbol n+1 on the right side may be performed as follows:
.function..times.'.function..times.'.function..times.'.function..times..t-
imes. ##EQU00006## In equation (9), the filtered CIRE {tilde over
(h)}.sub.M(n) for the current OFDM symbol period is determined
based on the initial CIREs for the previous, current, and next OFDM
symbol periods.
In an embodiment, the filtering for a "left edge" OFDM symbol n
that is bordered by only OFDM symbol n+1 on the right side may be
performed as follows:
.function..times.'.function..times.'.function..times..times.
##EQU00007## In equation (10), the filtered CIRE {tilde over
(h)}.sub.M(n) for the current OFDM symbol period is determined
based on the initial CIREs for the current and next OFDM symbol
periods.
In an embodiment, the filtering for a "right edge" OFDM symbol n
that is bordered by only OFDM symbol n-1 on the left side may be
performed as follows:
.function..times.'.function..times.'.function..times..times.
##EQU00008## In equation (11), the filtered CIRE {tilde over
(h)}.sub.M(n) for the current OFDM symbol period is determined
based on the initial CIREs for the previous and current OFDM symbol
periods.
In general, the time filtering of the initial CIREs may be
performed across any number of past and/or future OFDM symbols.
Furthermore, the time filtering may be performed with a finite
impulse response (FIR) filter, e.g., as shown in equations (9)
through (11), an infinite impulse response (IIR) filter, or some
other types of filter. The filtering may also be adaptive, e.g.,
adjusted based on the velocity of the receiver, the rate of changes
in the channel conditions, the operating SNR, and so on.
The receiver may perform thresholding on the filtered CIRE {tilde
over (h)}.sub.M(n) to obtain a final CIRE h.sub.M(n). The filtered
CIRE {tilde over (h)}.sub.M(n) contains M channel taps {tilde over
(h)}.sub.1(n) through {tilde over (h)}.sub.M(n). Each channel tap
{tilde over (h)}.sub.m(n), for m=1, . . . , M, has a complex gain
determined by the wireless channel. The thresholding retains
channel taps with sufficient energy and discards weak channel
taps.
In an aspect, the thresholding is performed in accordance with a
threshold parameter and a threshold. To derive the threshold, the
average channel energy for the M channel taps in {tilde over
(h)}.sub.M(n) may be computed as follows:
.function..times..function..times..times. ##EQU00009## where {tilde
over (h)}.sub.m(n) is the m-th channel tap in {tilde over
(h)}.sub.M(n), and
E.sub.avg(n) is the average channel energy for OFDM symbol period
n.
In an embodiment, the threshold is defined based on the average
channel energy and the threshold parameter, as follows:
T.sub.h(n)=PE.sub.avg(n), Eq (13) where P is the threshold
parameter and T.sub.h(n) is the threshold for OFDM symbol period n.
The threshold parameter may also be called a threshold constant, a
scaling factor, and so on. The threshold may also be defined as
T.sub.h(n)=P.sub.iE.sub.total(n), where E.sub.total(n) is the total
channel energy and P.sub.i=P/M is a revised threshold
parameter.
In general, the threshold T.sub.h(n) may be a function of any
quantities. The threshold may be a function of the average channel
energy and the threshold parameter, e.g., as shown in equation
(13). Alternatively or additionally, the threshold may be a
function of the noise energy, the energy of some number of weak
channel taps, the strongest channel tap energy, and so on.
The receiver may perform thresholding of the filtered CIRE, as
follows:
.function..function..times..times..function..gtoreq..function..times..tim-
es..times..times..times..times..times. ##EQU00010## where
h.sub.m(n) is the m-th channel tap in h.sub.M(n). In the embodiment
shown in equation (14), the thresholding is performed individually
for each of the M channel taps in {tilde over (h)}.sub.M(n). The
energy of each filtered channel tap {tilde over (h)}.sub.m(n) is
computed and compared against the threshold T.sub.h(n). The final
channel tap h.sub.m(n) is set to the filtered channel tap {tilde
over (h)}.sub.m(n) if the energy meets or exceeds the threshold
T.sub.h(n) and is set to zero otherwise.
FIG. 4 illustrates thresholding for an exemplary channel impulse
response estimate 400. The energies of the M filtered channel taps
are shown by vertical lines with different heights at tap indices 1
through M. The threshold T.sub.h(n) is shown by a dash line 410.
Channel taps with energies above line 410 are retained, and weak
channel taps with energies below line 410 are zeroed out. As can be
seen from FIG. 4, raising the threshold and line 410 (by increasing
the threshold parameter) may result in more channel taps being
zeroed out. Conversely, lowering the threshold and line 410 (by
decreasing the threshold parameter) may result in more channel taps
being retained.
FIG. 4 and the description above are for one thresholding
embodiment. The thresholding may also be performed in other
manners. For example, the channel taps may be ranked from strongest
to weakest. Channel taps may then be zeroed out, one channel tap at
a time starting with the weakest channel tap, until some percentage
of the total energy is discarded, some percentage or number of
channel taps is zeroed out, and so on. The percentage may be
determined by the threshold parameter P.
The thresholding may be performed on the channel taps {tilde over
(h)}.sub.m(n) in the filtered CIRE, as described above. The
thresholding may also be performed on the channel taps h'.sub.m(n)
in the initial CIRE, without filtering.
The receiver may use the final CIRE h.sub.M(n) for various purposes
such as data detection, log-likelihood ratio (LLR) computation, and
so on. For example, the receiver may derive a final channel
frequency response estimate H.sub.K(n) for all K total subbands
based on the final CIRE h.sub.M(n) with M channel taps. The
receiver may then perform equalization or matched filtering on the
received data symbols in Y.sub.K(n) with the final channel
frequency response estimate H.sub.K(n) to obtain data symbol
estimates {circumflex over (X)}.sub.K(n). The receiver may also use
H.sub.K(n) to compute LLRs for the bits of the data symbol
estimates.
Computer simulations were performed for the exemplary OFDM system
shown in FIGS. 2 and 3 with K=1024, G=136, U=888, M=128, P=111, and
C=108. Six different operating scenarios corresponding to two
channel models and three combinations of code rate and modulation
scheme were simulated. For each simulated operating scenario,
performance was characterized for different threshold parameter
values. The simulations indicate that the threshold parameter has a
large impact on both the quality of the channel estimate and
performance. Table 1 gives the threshold parameter values that
provide the best performance for the six operating scenarios
simulated.
TABLE-US-00001 TABLE 1 Threshold Parameter Values with Best
Performance Dual Cluster Vehicle A Pedestrian B Coding and
Modulation VEHA (120 Kmph) PEDB (120 Kmph) QPSK, Rate 0.55 P = 0.75
P = 1.00 16-QAM, Rate 0.41 P = 0.50 P = 0.75 16-QAM, Rate 0.55 P =
0.25 P = 0.50
VEHA and PEDB are two channel profile models that are well known in
the art. A channel profile is a statistical model for a channel
impulse response and is indicative of how the communication channel
looks like in the time domain. A channel profile is dependent on
speed and environment.
The results in Table 1 are obtained with large data block sizes and
Turbo coding on data blocks sent across 12 OFDM symbols in four
slots. A data block may also be called a packet, a frame, and so
on. For the dual cluster VEHA model, the first cluster starts at 0
.mu.s, the second cluster starts at 10 .mu.s, both clusters have
equal power, and the transmit pulse is a full sinc function.
Each combination of code rate and modulation scheme requires a
certain minimum SNR in order to achieve a target block error rate
(BLER), e.g., 1% BLER. In Table 1, the required SNR for rate 0.55
with QPSK is lower than the required SNR for rate 0.41 with 16-QAM,
which is lower than the required SNR for rate 0.55 with 16-QAM. For
a given modulation scheme, a higher code rate corresponds to a
higher required SNR. For a given code rate, a higher order
modulation scheme corresponds to a higher required SNR. Table 1
indicates that a higher threshold parameter value may provide
better performance at lower SNRs for a given channel profile.
Table 1 gives results for some exemplary operating scenarios. In
general, an operating scenario may be characterized by a channel
profile, an operating SNR, a coding and modulation scheme, some
other parameters, any one of the parameters, or any combination of
the parameters. Various operating scenarios may be simulated to
determine the threshold parameter values that provide the best
performance for these operating scenarios. Different results may be
obtained with different system parameters, channel profile models,
and/or assumptions.
The proper value to use for the threshold parameter P may be
determined in various manners. In one embodiment, threshold
parameter values that provide good performance for various
operating scenarios may be determined by computer simulations,
empirical measurements, and so on, and may be stored in a look-up
table. Thereafter, the current operating scenario for the receiver
may be ascertained, e.g., based on the channel profile, the coding
and modulation scheme, and/or other parameters applicable for the
receiver. The threshold parameter value corresponding to the
current operating scenario is retrieved from the look-up table and
used for channel estimation.
In another embodiment, the threshold parameter value P is selected
based on an expected operating SNR. The operating SNR may be
estimated based on received pilot symbols and/or received data
symbols. In general, a smaller threshold parameter value may be
used for a higher SNR, and a larger threshold parameter value may
be used for a lower SNR.
In yet another embodiment, the threshold parameter value P is
selected based on the number of channel taps in the CIRE. The
number of channel taps may be determined by the number of subbands
used for pilot transmission, the manner in which channel estimation
is performed at the receiver, and possibly other factors.
In yet another embodiment, the threshold parameter value P is
determined based on a high-quality channel estimate. The receiver
may obtain the high-quality channel estimate, e.g., based on a TDM
pilot or via some other means. The channel profile for the receiver
may be ascertained based on the high-quality channel estimate, and
a threshold parameter value may be selected based on the channel
profile.
In an embodiment, a new threshold parameter value is selected
whenever a higher quality channel estimate is desirable. For
example, a new threshold parameter value may be selected if a
packet is decoded in error. The new threshold parameter value may
be obtained as follows: P.sub.new=P.sub.old+.DELTA.P, or Eq (15)
P.sub.new=P.sub.old-.DELTA.P, where P.sub.old is the old/current
threshold parameter value,
P.sub.new is the new threshold parameter value, and
.DELTA.P is a step size, which may be set to 0.25 or some other
value.
A new channel estimate may be derived based on the new threshold
parameter value and used to recover the packet. If the packet is
still decoded in error with the new channel estimate, then another
threshold parameter value may be selected and used to derive
another channel estimate, which may then be used to recover the
packet. In general, any number of channel estimates may be derived
with different threshold parameter values. New threshold parameter
values may be selected from both sides of the original threshold
parameter value in alternate manner. For example, the new threshold
parameter value may be set to P.sub.old+.DELTA.P, then to
P.sub.old-.DELTA.P, then to P.sub.old+2.DELTA.P, then to
P.sub.old-2.DELTA.P, and so on. A new threshold parameter value may
be selected and used until the packet is decoded correctly, the
maximum number of values has been tried, or some other termination
condition is encountered. If the packet is decoded correctly, then
the threshold parameter value that results in successful decoding
may be used for subsequent packets. Selection of a new threshold
parameter value may also be triggered by other events besides
packet error.
FIG. 5 shows a block diagram of an embodiment of channel
estimator/processor 170 in FIG. 1. Within channel
estimator/processor 170, a pilot demodulator (Demod) 512 removes
the modulation on received pilot symbols and also provides zero
symbols for unused pilot subbands. A CIRE processor 514 derives an
initial CIRE for the current symbol period based on the output of
pilot demodulator 512. CIRE processor 514 may derive the initial
CIRE based on the least-squares technique shown in equation (6),
the MMSE technique shown in equation (7), the robust MMSE technique
shown in equation (8), or some other technique. A filter 516
filters the initial CIREs for different symbol periods, e.g., as
shown in equations (9), (10) and (11), and provides a filtered CIRE
for the current symbol period.
Controller 190 ascertains the current operating scenario and
selects an appropriate threshold parameter value for the current
operating scenario. Memory 192 may store a look-up table (LUT) of
different threshold parameter values for different operating
scenarios. A threshold computation unit 520 derives the threshold
T.sub.h(n) for the current symbol period based on the filtered CIRE
and the threshold parameter value, e.g., as shown in equations (12)
and (13). A unit 518 performs thresholding on the channel taps of
the filtered CIRE based on the threshold from unit 520 and provides
a final CIRE for the current symbol period. An FFT unit 522 may
derive a channel frequency response estimate, if needed, based on
the final CIRE.
FIG. 6 shows an embodiment of a process 600 for performing channel
estimation with thresholding. An initial CIRE is derived for each
symbol period with pilot transmission (block 612). The initial CIRE
may be derived based on received pilot symbols for used pilot
subbands and zero symbols for zeroed-out pilot subbands. The
initial CIRE may also be derived based on the least-squares, MMSE,
robust MMSE, zero-forcing, or some other technique. A filtered CIRE
is derived for the current symbol period by filtering the initial
CIREs for the current, prior and/or future symbol periods (block
614). A first CIRE having multiple channel taps may be set to the
initial CIRE or the filtered CIRE for the current symbol period
(block 616).
A threshold parameter value is selected based on at least one
criterion (block 618). For example, the threshold parameter value
may be selected based on the channel profile, the operating SNR,
the number of channel taps, and so on. A threshold is derived based
on the first CIRE and the threshold parameter value (block 620). In
an embodiment, the average energy of the channel taps in the first
CIRE is determined, and the threshold is derived based on the
average energy and the threshold parameter value. A second CIRE is
derived by zeroing out selected ones of the channel taps in the
first CIRE based on the threshold (block 622). In an embodiment,
channel taps with energy less than the threshold are zeroed out to
obtain the second CIRE. The second CIRE may also be derived by
performing thresholding on the channel taps in other manners.
A determination is then made whether an improved channel estimate
is desired (block 624). An improved channel estimate may be desired
if a packet is decoded in error. If the answer is `Yes` for block
624 and if a termination condition is not encountered in block 626,
then a new threshold parameter value is selected, e.g., by varying
the current threshold parameter value by .DELTA.P (block 628). The
process then returns to block 620 to (1) determine a new threshold
based on the new threshold parameter value and (2) derive a new
second CIRE by zeroing selected ones of the channel taps of the
first CIRE based on the new threshold. Blocks 620 through 628 may
be performed any number of times until a termination condition is
encountered. If an improved channel estimate is not desired, as
determined in block 624, or if a termination condition is
encountered, as determined in block 626, then the process
terminates.
Those of skill in the art would understand that information and
signals may be represented using any of a variety of different
technologies and techniques. For example, data, instructions,
commands, information, signals, bits, symbols, and chips that may
be referenced throughout the above description may be represented
by voltages, currents, electromagnetic waves, magnetic fields or
particles, optical fields or particles, or any combination
thereof.
Those of skill would further appreciate that the various
illustrative logical blocks, modules, circuits, and algorithm steps
described in connection with the embodiments disclosed herein may
be implemented as electronic hardware, computer software, or
combinations of both. To clearly illustrate this interchangeability
of hardware and software, various illustrative components, blocks,
modules, circuits, and steps have been described above generally in
terms of their functionality. Whether such functionality is
implemented as hardware or software depends upon the particular
application and design constraints imposed on the overall system.
Skilled artisans may implement the described functionality in
varying ways for each particular application, but such
implementation decisions should not be interpreted as causing a
departure from the scope of the present invention.
The various illustrative logical blocks, modules, and circuits
described in connection with the embodiments disclosed herein may
be implemented or performed with a general purpose processor, a
digital signal processor (DSP), an application specific integrated
circuit (ASIC), a field programmable gate array (FPGA) or other
programmable logic device, discrete gate or transistor logic,
discrete hardware components, or any combination thereof designed
to perform the functions described herein. A general purpose
processor may be a microprocessor, but in the alternative, the
processor may be any conventional processor, controller,
microcontroller, or state machine. A processor may also be
implemented as a combination of computing devices, e.g., a
combination of a DSP and a microprocessor, a plurality of
microprocessors, one or more microprocessors in conjunction with a
DSP core, or any other such configuration.
The steps of a method or algorithm described in connection with the
embodiments disclosed herein may be embodied directly in hardware,
in a software module executed by a processor, or in a combination
of the two. A software module may reside in RAM memory, flash
memory, ROM memory, EPROM memory, EEPROM memory, registers, hard
disk, a removable disk, a CD-ROM, or any other form of storage
medium known in the art. An exemplary storage medium is coupled to
the processor such that the processor can read information from,
and write information to, the storage medium. In the alternative,
the storage medium may be integral to the processor. The processor
and the storage medium may reside in an ASIC. The ASIC may reside
in a user terminal. In the alternative, the processor and the
storage medium may reside as discrete components in a user
terminal.
The previous description of the disclosed embodiments is provided
to enable any person skilled in the art to make or use the present
invention. Various modifications to these embodiments will be
readily apparent to those skilled in the art, and the generic
principles defined herein may be applied to other embodiments
without departing from the spirit or scope of the invention. Thus,
the present invention is not intended to be limited to the
embodiments shown herein but is to be accorded the widest scope
consistent with the principles and novel features disclosed
herein.
* * * * *